I interviewed 100 writers about AI - here's what I learnt

Writing, like everything, is undergoing a massive transformation with the rise of generative AI. However, the writing community’s response is far from monolithic. After speaking with 100 writers - bloggers, essayists, academics, authors, etc. - from various backgrounds, expertise, and focuses, it’s clear that this future is as emergent as any today, forming live under our feet.

From the many conversations I’ve had about our current situation, patterns have started to emerge and form a picture of how an AI-writing future might unfold. Patterns I’ve tried to distill below. Whilst there’s still - and may always be - division around these tools, I have no doubt we are at the cusp of a real change to how we think about writing as a practice, an act, and a livelihood.

The Defining Philosophy

When asking this question, one of the biggest drivers of divergence comes down to a philosophical question (one I've written a bit about before). Although it’s one of multiple contributing factors, I think it deserves unique attention since it really does feel like the cleanest proxy we have for a simple answer.

Are ‘thinking’ and ‘language’ intrinsically linked, or separate acts altogether?

This philosophical stance seems to strongly correlate with how writers approach AI in their work:

• Outlook 1: Writing is an important medium for thinking. These writers believe the act of writing is inseparable from the thought process itself - writing is a means to gain clarity and understanding on a topic. Therefore, the integration of AI into the writing process is really an integration into the thinking process. Because of this, decisions around creativity, authenticity, and agency must be made accordingly. 

“writing is the main part of thinking… you need to write to think deeply… it helps you get more clarity” (Post-grad and academic writer)

• Outlook 2: Writing is simply a means to communicate thinking: These people view writing simply as a mechanism for expressing pre-formed thoughts, rather than a process to develop thinking itself. They therefore understand AI’s role in terms of practical value rather than focusing on more philosophical implications - supporting the act of putting the right words on paper, in the clearest and most engaging way possible.

"If I can just talk, take my brain-dump, and publish at the cadence I want, well, good enough for me. I don't give a shit if it's written by an AI." (Entrepreneur and blogger)

This philosophical divide doesn't just influence opinions - it helps predict behaviours. Writers who see thinking and writing as deeply interconnected approach AI with careful consideration, while those who view them as distinct acts tend to be more open to more experimentation. I think many reading this will see themselves in one of these camps - and if you’re not sure, but the idea of AI doing the main bulk of your writing gives you a slight icky feeling in your stomach, you’re probably in group 1.

Other Emergent Factors

Alongside the philosophical foundations, there are more tangible and practical factors that correlate with people’s outlook.

Purpose/incentive 

• Writing for the sake of ideas: Some write for clarity or to develop a deep expertise, whether they publish or not. They write to think and therefore have less interest in using AI for this process unless it can actively improve their ‘thinking’ - often exploring a relationship I explored in a different piece.

• Writing for the sake of return: Some write for money, community, following, or other returns-based reasons. These people are generally more open to AI in their process, usually as a time saving measure - caring more about publishing something ‘good enough’, than publishing something profound.

Confidence in writing:

• High confidence: Confident writers typically have a tone, voice, and practice already, so see less need for assistance. They might be open to AI supporting on certain elements, but want less intrusive AI features, rather to pick and choose exactly where it comes in.

• Low confidence: Less confident writers often appreciate more extensive assistance and guidance, hoping for a tool that can support with best practice to get them just writing more and feel confident in hitting publish at the end. This might even include doing most of the writing.

Professional vs Hobbyist

• Professional: Professional writers tend to approach AI with existential concerns about their craft and livelihood - they’re more likely to have made a name for themselves through writing, so perhaps fear losing their place in society, livelihood, or something more profound.

• Hobbyist: Hobbyists generally experiment more freely without career implications and are more likely to want support in fast progression or publishing regularly, so are more open to AI.

Time Constraints:

• Time rich: Those with more time often aim to develop their skills alongside AI assistance, rather than as a replacement. They might want to experiment with AI as a thinking aid, but since they have the time to do writing (because it’s their main role/source of income or they have time for other reasons), they have less of a deep desire for efficiency gains.

• Time poor: Time-poor individuals gravitate toward deeper AI integration for efficiency, since for them writing is a ‘nice to have’ amongst a rammed schedule or something they’re currently experiencing a high opportunity-cost to pursue. They therefore are more open to more automation - as busy people generally are with many parts of life.

Of course, the above are all interlinked and overlapping in many ways, with people being at varying levels of the spectrum across each and some cause and effect definitely present between these factors.

Emerging Archetypes

From these emergent patterns, we can see two overarching archetypes:

Most likely to adopt AI = time poor hobbyist writers with relatively low confidence, who write for the sake or gaining something external (money, a following, etc.)

Least likely to adopt AI = professional writers who are time rich (for writing, at least), and therefore have high confidence and largely write for the sake of pursuing ideas that interest them.

However, life is rarely as cut and dry and in reality we have closer to five distinct archetypes emerging from my conversations:

The Puritan: "Writing is sacred; AI cheapens the craft."

• Views writing as a deeply personal, almost spiritual practice

• Concerned about losing the cognitive benefits of wrestling with language

• Sees AI as a threat to authentic expression

• Often has a strong identity tied to being a writer (even if it’s not their job title)

The Curious: "I'm intrigued, but cautious."

• Enjoys writing but isn't ideologically opposed to assistance

• Experiments in low-stakes scenarios

• Actively questions which aspects of writing they're willing to delegate

• Often toggles between using and not using AI to compare experiences

The Pragmatist: "This is happening; I need to adapt."

• Recognises AI as ‘inevitable’ in the writing landscape

• Actively seeks the optimal human-AI collaboration model for themselves

• Often uses AI for specific sub-tasks while maintaining control over others

• Focused on maintaining quality while improving efficiency and exploration

The Optimist: "AI enhances rather than replaces."

• Excited about AI's potential to elevate their writing

• Views AI as a collaborator that handles mechanics whilst they focus on creativity

• Experiments broadly across their workflow

• Often reports unexpected creative breakthroughs from AI interactions

The Accelerationist: "Just get my ideas out there faster."

• Values communication over craftsmanship

• Embraces maximum AI involvement

• Focused primarily on output and reach

• Often sees traditional writing process as an unnecessary bottleneck to sharing their ideas

The Adoption Trajectory

Going into these conversations, I braced myself for waves of skepticism about AI and writing and whilst - yes - there is plenty of that, something genuinely surprised me: the adoption curve is already well and truly on the up and a crossing of the chasm feels much closer than I originally expected. Even those most puritan are exploring AI in this practice, and whilst they may grumble about its pitfalls, many continually come back to these tools. Perhaps this is just morbid curiosity or an attempt to feel some superiority over these tools whilst they’re still in infancy, but I think it’s something more. I think it’s the purest signal of a change that’s here to stay.

The gateway drug for many is research (e.g. Deep Research or Perplexity) and editorial review (e.g. asking for notes to help with structure or arguments). Often they are more attracted to the ‘thought partner’ possibilities of AI, rather than its ability to put words to page for them, but I think the line between the two continually grows thinner as inertia kicks in or normalisation of use becomes the every day.

Some fight AI more than others, but it seems almost everyone I’ve spoken to is experimenting more and more. If this trajectory holds, it seems inevitable that most writers will soon incorporate AI in at least one phase of their work and it will become a vital part of our sense-making abilities. Whilst there will always be holdouts who refuse these tools on principal or a genuine lack of interest, I've become increasingly convinced that the foundation of writing is transforming beneath our feet, faster and more profoundly than many might realise.